VOL.
5,
NO.
5,
M A Y
1966
Yasunobu, K. T. (1965), in Non-Heme Iron Proteins: Role in Energy Conversion, San Pietro, A., Ed., Yellow Springs, Ohio, Antioch. Tanaka, M., Nakashima, T., Benson, A. M., Mower, H. F., and Yasunobu, K. T. (1964a), Biochem. Biophys. Res. Cammun. 16, 422.
Tanaka, M., Nakashima, T., Mower, H. F., and Yasunobu, K. T. (1964b), Arch. Biochem. Biophys. 105,570.
Valentine, R. C. (1964), Bacteriol. Rev. 28, 497. Valentine, R. C., Mortenson, L. E., and Carnahan, J. E. (1963), J . Biol. Chem. 238, 1141.
Purification and Physical Characterization of Tyrosyl Ribonucleic Acid Synthetases from Escherichia coli and Bacillus subtilis" Richard Calendar and Paul Berg
Tyrosyl ribonucleic acid synthetases from Escherichia coli and Bacillus subtilis have been purified to near homogeneity. These enzymes have molecular weights of 95,000 and 88,000, respectively. They can be separated electrophoretically on starch or polyacrylamide gels. The E. coli enzyme contains 14 halfcystine residues/mole while the B. subtilis enzyme has ABSTRACT:
A
minoacyl RNA1 synthetases are one of a few types of known proteins which can discriminate between structurally similar nucleic acid molecules (Berg, 1961). How each aminoacyl RNA synthetase distinguishes the correct t-RNA from all others is not known; moreover, virtually nothing is known about the type of interactions which occur between the protein and the polynucleotide chain during the formation of an aminoacyl RNA. Several examples are known in which aminoacyl RNA synthetases from one species fail to utilize the corresponding t-RNA's from another species; complete cross-reaction between enzymes and t-RNA's from different species is also well documented (Benzer and Weisblum, 1961; Yamane and Sueoka, 1963; Doctor and Mudd, 1963). Quite possibly, studies of one of the aminoacyl RNA synthetases, isolated from species which do and do not cross-react with their respective t-RNA's, could shed some light on the nature of the interaction between the enzyme and the t-RNA. Our first approach was to compare two enzymes from species which show complete cross-reaction with their corresponding t-RNA's (Calendar and Berg, 1966). Later studies will extend the comparison to the same
* From the Department of Biochemistry, Stanford University School of Medicine, Palo Alto, California. Receiwd Februarj' 15, 1966. This work was supported in part by teaching and research grants from the National Institutes of Health. 1 Abbreviations used: RNA, ribonucleic acid; t-RNA, transfer R N A ; ATP, adenosine triphosphate.
CHARACTERIZATION
2-3. Antiserum to the E. coli enzyme neutralizes and precipitates the E. coli enzyme but does not cross-react with the B. subtilis enzyme. Antiserum to the B. subtilis enzyme neutralizes and precipitates the B. subtilis enzyme and also shows partial cross-reaction with the E. coli enzyme.
enzyme from species which fail to react with the heterologous t-RNA's. In this paper we report the isolation and a comparison of several physical and chemical properties of tyrosyl RNA synthetases from Escherichia coli and Bacillus subtilis, two enzymes which cross-react completely with their heterologous t-RNA's. Although the two enzymes have similar molecular weights, they are readily distinguishable by their electrophoretic mobility, amino acid composition. and immunological reaction. In the next paper (Calendar and Berg, 1966) the catalytic properties of the two enzymes are described. Experimental Section Muterials. L-Tyrosine was purchased from California Corp. for Biochemical Research and gave only a single peak when chromatographed on an amino acid analyzer (Spackman et al., 1958). ATP and glutathione were purchased from Sigma Chemical Corp. 32P-Labeled pyrophosphate ( 32PPi) was synthesized from 32P-labeledorthophosphate (Oak Ridge National Laboratories) as described by Bergmann et al. (1961). Crystalline bovine plasma albumin was purchased from Armour Pharmaceutical Co., Kankakee, Ill. Sucrose, Baker analyzed grade, was recrystallized from ethanol and filtered through a 0.45 1.1 Millipore filter before use in sucrose-gradient centrifugation. Fluorochemical FC43 was obtained
OF TYROSYL
R I B O N U C L t l C ACID
168 1
S Y N T H E T A S E
BIOCHEMISTRY
1682
from Beckman and Co., Palo Alto, Calif. Superbrite glass beads, 200-p average diameter, were from Minnesota Mining and Manufacturing Co., and were washed before use with 1 N HCl, then with water. Whatman phosphocellulose P11 was obtained from BKH Division of Van Waters and Rogers Corp., San Francisco, Calif., and DEAE-cellulose came from Brown and Co., Berlin, N. H., and from Bio-Rad Laboratories, Richmond, Calif. Phosphocellulose and DEAE-cellulose were prepared for use by removal of the fines through repeated settling, suspension in 0.25 M NaOH at room temperature for 10 min, and finally washing with water on a Biichner funnel. Both resins were stored in 0.1 M KHzPOl to maintain the pH below 5. Hydroxylapatite was obtained from Clarkson Chemical Corp., Williamsport, Pa., alumina Cy gel was from Sigma Chemical Corp., St. Louis, Mo., and was also prepared according to Willstatter and Kraut (1923). Hydrolyzed starch was purchased from Connaught Medical Research Laboratories, Toronto, Canada; acrylamide and methylene bisacrylamide were from Eastman Chemical Laboratories, Rochester, N. Y . ,and were recrystallized from acetone before use. Virgin female rabbits were used for antibody production. Freund's adjuvant and Noble agar were the products of Difco Laboratories, Detroit, Mich. Normal rabbit y-globulin was obtained from Nutritional Biochemicals Corp., Cleveland, Ohio. and was dialyzed cs. distilled water before use. E . coli B or B/l (a T1 phage-resistant mutant isolated from prototrophic E. coli B) and B. subtilis SB19 (a prototrophic strain obtained from Dr. Esther Lederberg and described by Nester et ul., 1963) were the starting materials for the enzyme isolation. E. coli was grown in large quantities in a 120-1. biogen apparatus (American Sterilizer Corp.) at 37 O with forced aeration. M56 medium as described by Weismeyer and Cohn (1960) was used with 1 glucose as carbon source. The cells were removed in late log phase, quick cooled to 5 O by passage through a cooling coil immersed in a - 15O ethylene glycol bath, centrifuged in a Spinco Model 170 continuous flow centrifuge at 29,00Og, and frozen in thin layers at -20". Cells stored in a closed container were satisfactory for at least 6 months. B. subtilis SB19 were grown in large quantities using the same procedure as described for E . coli B, except that the medium was Spizizen (1958) minimal + 1 % glucose, plus 0.1 casein amino acids (acid hydrolyzed, Nutritional Biochemicals) and 40 pg/ml of L-tryptophan (B grade, Calbiochem). Isoleucyl RNA synthetase was a gift from Dr. Anne N. Baldwin. Methods. ENZYME ASSAY. The activity of the enzymes were measured by the L-tyrosine-dependent exchange of ATP with 32PP,. The reaction mixture (1 ml) contained: 100 mM sodium cacodylate buffer, pH 7.0, 5 mM magnesium chloride (MgCI2), 2 mM adenosine triphosphate (ATP), 2 mM sodium pyrophosphate (10e106 cpm/pmole), 10 mM 2mercaptoethano1, 10 mM potassium fluoride (KF) (for crude fractions only), 0.1 mg of bovine plasma albumin, 2 mM L-tyrosine, and enzyme (0.01 to 0.1 unit) in a solution usually con-
RICHARD
C A L E N D A R
A N D
P A U L
BERG
taining 50 mM potassium phosphate, pH 6.5, 10 mM 2mercaptoethanol, and 0.1 mg/ml of bovine plasma albumin. After 15 min at 37", the reaction was stopped by the addition of 0.7 ml of cold 0.4 M sodium pyrophosphate in 15% perchloric acid. The ATP was adsorbed to charcoal by the addition of 0.1 ml of a 15 suspension of acid-washed Norit A. The charcoal suspension was filtered through a GFjC glass filter, and the filter was washed five times with 5-ml portions of cold distilled water. The filters were dried under a heat lamp and counted in an end-window counter. The values were corrected for a blank which was run without tyrosine ; no self-absorption correction was made. One unit of enzyme activity is defined as that amount incorporating 1 pmole of 32PP,into ATP in 15 min, under the conditions described, and the specific activity is defined as enzyme units per milligram of protein. The exchange of 3*PP, with ATP was proportional to the amount of enzyme added (up to 0.15 unit) and to time for periods up to 20 min. PROTEINASSAY. Protein was determined by the method of Lowry et al. (1951). In crude fractions, protein was precipitated with cold 5 % trichloroacetic acid before assay. ELECTROPHORESIS. Starch gel electrophoresis was performed as described by Smithies (1959). Gels (15 %) (w/v) were prepared in 20 mM potassium phosphate, pH 8.0, containing 10 mM 2-mercaptoethanol. Each slit was filled with 60 p1 of enzyme solution (360 pg) and after 36 hr at 200 v the gel was sliced in half horizontally, and one-half was stained with amido black, while the other half was cut transversely into strips, eluted with 100 mM potassium phosphate, pH 7.0, containing 10 mM 2-mercaptoethanol. and assayed for enzyme activity. Electrophoresis on polyacrylamide gels was performed as described by Jovin el al. (1964). Separating gels were prepared containing 6 % acrylamide and 0.12 % bisacrylamide, while stacking gels were 2.5 acrylamide and 0.6 % bisacrylamide. The Tris buffer system was used. Two milliamperes per gel was applied at 5 O until a brom phenol blue marker had reached the end of the gels (ca. 3 hr), and then parallel gels were stained or sliced, eluted, and assayed for enzymatic activity. SUCROSEGRADIENT CENTRIFUGATION was carried out after Martin and Ames (1961). Gradients (4.8 ml) from 20 to 5 sucrose in 20 mM potassium phosphate, pH 6.5, 100 mM KCI, 10 mM 2-mercaptoethanol, were prepared. Purified enzyme (1 mg) in the same buffer was layered on top of the gradient. Centrifugation was at 39,000 rpm in a Spinco Model L ultracentrifuge for 16 hr at 3". Two or three-drop fractions were collected from the bottom of the tube and assayed for protein content and enzyme activity. IMMUNOLOGICAL TECHNIQUES. Rabbits were immunized by a primary subcutaneous injection of 100 pg (1 ml) of enzyme in an equal volume of adjuvant, followed by a secondary intravenous injection 6 weeks later with 10 pg of enzyme in isotonic saline. Blood
VOL.
5,
NO.
5,
MAY
1966
was collected by heart puncture and allowed to clot, and the serum was collected arter centrifugation of the clot. y-Globulin was purified by precipitation of whole serum with ammonium sulfate to 50% saturation at 5". The precipitate was dissolved in 0.016 M potassium phosphate, pH 6.5, dialyzed 1;s.the same buffer, and chromatographed on a column of DEAE-cellulose at the same buffer concentration. The material not adhering to the column (y-globulin) was concentrated by pervaporation (Richardson and Kornberg, 1964). Studies of antigen-antibody precipitin reactions were carried out by double-diffusion in agar (Ouchterlony, 1948), using gels of 2 Noble agar in 50 mM sodium barbital buffer, pH 8.6. Acid fixing of the precipitin lines with 5 % acetic acid was done after the gel had been washed for a day with 150 mM NaCl. We are indebted to Dr. Jack Remington and Mr. Hilbert Morales for advice and aid in the immunological work. MoIeculur Weight Studies. Sedimentation studies were done in a Spinco Model E analytical ultracentrifuge, equipped with temperature control and Schlieren optics, using both titanium and black anodized aluminum rotors. A 12-mm 2.5" Epon double-sector centerpiece with quartz windows was used for all studies, except that a 2.5 " single-sector centerpiece was used for determinations of the sedimentation coefficient with the B. subtitis tyrosyl RNA synthetase. Diffusion coefficients were determined in boundary-forming runs, using a capillary-type, synthetic-boundary, double-sector cell. All centrifugal studies were performed in solutions containing 20 mM potassium phosphate, pH 6.5, 100 mM KCI, and 3 mM 2-mercaptoethanol. The rotor temperature was maintained at a constant value, which for most runs was between 5 and 10". The rotor speed for sedimentation velocity runs was 50,740 rpm, while speeds of 8225-19,160 rpm were used for sedimentation equilibrium runs. Photographic plates were measured on a Gaertner M2001 RS two-dimensional microcomparator . Calculations of s20,w and D 2 ~ , were w according to Schachman (1957). Molecular weight was calculated from the equilibrium data by method I and method I1 of Van Holde and Baldwin (1958), and by plotting In c (c = protein concentration) us. r2 (radius) after Schachman (1957). Method I utilized AC/COto give a weightaverage molecular weight, while method I1 utilizes a graph of (l/r)(dc/c&)US. c - c,, which should be a straight line for homogeneous proteins, and whose slope defines an M z . The plot of In c cs. r2 gives the molecular weight ( M * ) if the protein is homogeneous. Molecular weights were also obtained from sl0,,,. and D?o.,~,by use of the Svedberg equation (Svedberg and Pedersen, 1940). AMINOACID ANALYSES. Protein (1 mg) was dialyzed c's. 100 mM KCI, and then cs. distilled water. This material was hydrolyzed in uucuo at 110" for 22 hr in 6 N HCI, with a crystal of phenol to protect tyrosine from oxidation (Sanger and Thompson, 1963). The resultant hydrolysate was analyzed on a Beckman amino acid analyzer, Model 120B (Spackman et al., 1958). Half-cysteine residues were determined as cysteic
CHARACTERIZATION
acid after performic acid oxidation according to Moore (1963), and tryptophan was determined by spectral analysis in 100 mM NaOH (Beaven and Holiday, 1952) on a Beckman PMQ I1 spectrophotometer. Since all hydrolyses were for 22 hr, no corrections could be made for destruction of amino acids or for incomplete hydrolysis. Results A . PuriJcution of Tyrosyl RNA Synthetuse from E . coli (a summary is given in Table I.) All operations
Purification of Tyrosyl RNA Synthetases from E. ~oli.5
TABLE I:
Fraction
Enzyme Sp Act Vol (units/ (units/ Yield (mu ml) mg) (%I
2225 I. Crude extract 11. Autolyzed extract 4225 111. Ammonium sulfate 324 IV. AluminaCrgel 750 V. DEAE-cellulose, 324 pH 6.5 VI. Hydroxylapatite 150 VII. DEAE-cellulose, 50 pH 8.0
32 15 142 39 60 154 142
1.5 2.1 6.5 24 57 294 910
100 89 65 41 27 32 10
a Seven preparations yielded material with final specific activities ranging from 500 to 910 units/mg, with an average of 660.
were carried out at 4" and all centrifugations were in the GSA rotor of a Sorvall centrifuge at 8000 rpm. I. PREPARATION OF EXTRACT. Cells (500 g) (wet weight), 500 ml of 25 mM K2HPO4,and 1 kg of glass beads were blended for 10 min at low speed in a stainless-steel Waring Blendor maintained below 5" by circulating ethylene glycol at - 15" through a cooling jacket. After 10 min, 1 1. of 25 mM KJ-IPO, was added, and blending was continued for 3 min. The supernatant fluid was decanted and the beads were washed twice with 500-ml aliquots of 25 mM K2HP04.The supernatant fluid and the two washes were pooled and centrifuged for 40 min, and the supernatant fluid (fraction I) was used in succeeding steps. 11. AUTOLYSIS. Fraction I was adjusted to 4.75 1. (cu. 10 mg of protein/ml) and the phosphate concentration was adjusted to 100 mM by addition of 80 g of solid K,HPO,. During 2 hr at 37" the acid-soluble A260rose to 70% of the total A260 with little loss in enzyme activity. The material was cooled to 5" and a precipitate was eliminated by centrifugation (fraction 11). 111. AMMONIUM SULFATE FRACTIONATION. Ammonium sulfate (34 g) was added for every 100 ml of fraction I1
OF
TYROSYL RIBONUCLEIC
ACID
1683
SYNTHETASE
UIOCHEMIS I K Y
TABLE 11:
Purification of Tyrosyl RNA Synthetase from B. subtilis:
Fraction I. 11. 111. IV. V. VI. VII. VIII.
Crude extract Autolyzed extract Ammonium sulfate, 60-75 Aumina Cy gel DEAE-cellulose, pH 6.0 Cellulose phosphate Hydroxylapatite DEAE-cellulose, pH 8.0
Vol (ml)
Enzyme (units/ml)
Sp Act (units/mg)
3380 3500 173 346 340 350 240 38
3.5 4.3 72 25 14 6 8 25
0.6 0 8 3.2 6.2 15 58 180 720
.
Yield
(z)
100 128 111 74 42 20 18 8
Four preparations yielded final fractions with specific activities of 440, 510, 500, and 720.
and after 5 min the suspension was centrifuged for 30 min. Additional ammonium sulfate (12.9 g/100 ml of fraction 11) was added to the supernatant fluid, and after 5 min the mixture was centrifuged for 30 min. The precipitate was dissolved in ca. 250 ml of 10 mM potassium succinate buffer, pH 6.0, containing 3 mM 2mercaptoethanol and the solution was dialyzed overnight us. two successive 30-1. aliquots of the same buffer. Any precipitate which appeared during dialysis was removed by centrifugation (fraction 111). From this point all solutions contained 3 mM 2-mercaptoethanol. IV. ALUMINA C , GEL FRACTIONATION. Alumina C y gel2 was added to fraction I11 (58 mg of dry weight gel/mg of protein) and the suspension was centrifuged immediately. The gel was washed twice with 375-m1 aliquots of 20 mM potassium phosphate buffer, pH 7.5, and then eluted twice with 375-ml aliquots of 110 mM potassium phosphate buffer, pH 7.5. The pooled supernatant fluids were fraction IV. V. DEAE-CELLULOSE FRACTIONATION. Fraction IV was diluted threefold with 3 mM 2-mercaptoethanol and applied to a column of DEAE-cellulose3 (20 cm X 3.1 cmz) which had been previously equilibrated with 50 mM potassium phosphate, pH 6.5. Elution was effected with a linear gradient between 50 and 300 mM potassium phosphate, pH 6.5, the total volume of the eluent being 1 1. Fractions (20 ml) were collected and fractions containing enzyme having a specific activity at least twice that of fraction IV were pooled (fraction V). VI. HYDROXYLAPATITE FRACTIONATION. Fraction V
1684
2 This particular enzyme preparation was made with C, gel prepared according to Willstatter and Kraut (1923); commercially obtained Cygel also gave good fractionation, but the buffer concentrations used for washing and elution vary somewhat from those given here, and should be checked for each preparation of gel. 3 With DEAE-cellulose purchased from Brown Paper Co. (0.9 mequiv/g) the enzyme was eluted at 120 m phosphate, while with DEAE-cellulose purchased from Bio-Rad Labs (0.8 mequiv/g) a higher phosphate concentration (180 mM) was needed to elute the enzyme.
RICHARD
CALENDAR
A N D
P A U L BERG
was diluted threefold with 3 mM 2-mercaptoethanol and applied to a column of hydroxylapatite (25 cm X 0.75 cmz) previously washed with 1 1. of 50 mM potassium phosphate, pH 6.5. The column was then washed with 200 ml of 90 mM potassium phosphate, pH 6.5, and the enzyme was eluted (1 ml/min, 2 psi of pressure) with a linear gradient from 90 to 250 mM potassium phosphate, pH 6.5, the total volume of the eluent being 400 ml. Fractions (7 ml) were collected and fractions containing enzyme with a specific activity at least twice that of fraction V were pooled (fraction VI). VII. SECOND DEAE-CELLULOSE FRACTIONATION. Fraction VI was diluted threefold with 3 mM 2-mercaptoethanol and applied to a column of DEAE-cellulose (30 cm X 0.75 cm2) previously equilibrated with 50 mM potassium phosphate, pH 8.0. The enzyme was eluted with a linear gradient between 50 and 300 mM potassium phosphate, pH 8.0, the total volume of the gradient being 500 ml. Fractions (10 ml) were collected, and the fractions of highest specific activity were pooled (fraction VII). For storage, fraction VI1 was concentrated to 1 ml by pressure dialysis DS. 20 m potassium phosphate, pH 6.5, 3 mM in 2-mercaptoethanol, or glutathione. Concentrated fractions containing mercaptoethanol were stable at 0" for several months. Fractions with glutathione could be frozen indefinitely in a liquid nitrogen refrigerator. Upon thawing, addition of 2-mercaptoethanol was necessary to restore full activity. Minor electrophoretic bands were present at this stage in the preparation and the following additional procedure was carried out on one sample to remove these contaminants. VIII. PREPARATIVE ELECTROPHORESIS. Fraction VI1 was subjected to preparative electrophoresis on polyacrylamide gel as described by Jovin et a f .(1964), using acrylamide-bisacrylamide (6-12 separation gel, acrylamide-bisacrylamide (2.5-0.6 %) concentration gel, and the Tris buffer system. The recovery of enzymatic activity was 7 5 z and there was a 25% increase in specific activity. B. Purification of Tyrosyl RNA Synthetase from B.
z)
VOL.
5,
NO.
5,
MAY
1966
, L l L L~~~~. 9
m,s.o":~
,
,n /m
1: Electrophoresis of purified tyrosyl RNA synthetases on starch gel (Smithies, 1959). B. subtilis enzyme fraction VI11 and E. coli enzyme fraction VI1 were used. FIGUIU!
subtilis SBI9 (see Table I1 for summary.) Tempera-
tures below 5" were maintained throughout, and centrifugations were as described in the previous purification. I. CRUDEEXTRACT. Crude extract (fraction I) was prepared as in the previous purification, except that 50 mM potassium phosphate was the buffer used, and the whole procedure was scaled up for 800 g of cells. 11. AUTOLYSIS.Fraction I was diluted to contain 5 mg of protein/ml and 100 mM potassium phosphate by the addition of potassium phosphate, pH 7.5. After incubating the extract for 150 min at 37', the amount of acid-soluble Also rose to 7 0 x of the total AZe0,with an increase of approximately 20% in enzyme activity (fraction 11). 111. AMMONIUM SULFATE FRACTIONATION. Ammonium sulfate (38 g) was addedjlw ml of fraction I1 and after 5 min the suspension was centrifuged for 20 min. An additional 11 g of ammonium sulfate was added to the supernatant fluid/initial 100 ml of fraction I1 and after 5 min the suspension was centrifuged for 20 min, the precipitate was dissolved in ca. 300 ml of 10 m M potassium succinate buffer, pH 6.0, containing 3 m M 2mercaptoethanol, and the solution was dialyzed overnight us. two successive 30-1. aliquots of the same buffer (fraction 111). From this point all solutions contained 3 m M 2-rnercaptoethanol.' I v . ALUMINA c, GEL FRACTIONATION. Alumina c, gels (50 mg of dry weight gel/mg of protein) was added to fraction 111 and the suspension was centrifuged immediately. The gel was washed twice with 250-ml aliquots of 20 mM potassium phosphate buffer, pH 7.0, then eluted twice with 250-ml aliquots of 75 m M potassium phosphate, pH 7.5 (fraction IV). V. DEAE-C~LLULOSE FRACTIONATION. Fraction IV was diluted twofold with 3 m M 2-mercaptoethanol and 4
There is at present no evidence that SH reagent is required
for the maintenance of enzyme activity. $The gel used was that purchased from Sigma Chemical Co.
C H A R A C T E R I Z A T IO N
FIGURE 2:
Electrophoresis of purified B. subtilis (fraction VIII) and E. coli (fraction VIII) enzymes on polyacrylamide gel (Jovin et al., 1964)
applied to a DEAEcellulose column (30 cm X 7.1 ems) which had been equilibrated previously with 40 mM potassium phosphate, pH 6.0. The enzyme was eluted with a linear gradient between 40 and 300 m M potassium phosphate, pH 6.0, the total volume of the gradient being 2 1. Fractions of 30 ml were collected and those having a specific activity at least twofold greater than fraction IV were pooled (fraction V). VI. PHOSPHCCELLULOSE FRACTIONATION.Fraction V was dialyzed us. two successive 30-1. portions of 10 mM succinate buffer, pH 6.0, and then applied to a column of phosphocellulose (30 cm X 7.1 cml) which had been equilibrated previously with 10 m M potassium succinate, pH 6.0.6 The column was washed with ca. 600 ml of the 10 m M succinate, pH 6.0, and then the enzyme was eluted with a linear gradient between 10 m M potassium phosphate, pH 6.0, and 200 m M potassium phosphate, pH 8.0, the total volume of the gradient being 1.5 1. The fractions, 25 ml each, containing enzyme activity were pooled (fraction VI). VII. HYDROXYLAPATITE FRACTIONATION. Fraction VI was diluted with 3 m M 2-mercaptoethanol to reduce the concentration of phosphate to ca. 50 m. The diluted fraction was adsorbed to a hydroxylapatite column (15 cm X 3.1 cmz) previously washed with 1 1. of 50 m potassium phosphate, pH 6.5. The enzyme was eluted with a linear gradient between 50 and 300 m M potassium phosphate, pH 6.5, the total volume of the eluent being 1 I. The fractions containing enzyme with a specific activity at least two times greater than fraction VI were pooled (fraction VII). VIII. s!XOND DEAE-CELLULOSE FRACTIONATION. FIXtion VI1 was diluted twofold with 3 m M 2-mercaptoethanol and passed over a DEAE-cellulose column (25 cm X 0.75 cm*) previously equilibrated with 50 mM potassium phosphate, pH 8.0. Enzyme was eluted
a It is important lo check the pH and conductivity of both the dialyzed enzyme and column wash fluid before carrying out this step, since the enzyme will not be quantitatively adsorbed at higher pH.
0F
'T Y R 0 S Y L R I U 0 N U C L E I C A C I 1)
SY N
1685
'r H E T A S E
BIOCHBMISTRY
FIGURE 3: Sucrose gradient centrifugation (Martin and Ames. 1961) of purified tyrosyl RNA synthetases: (a) E. coli fraction VII; enzyme recovery 87 %; protein recovery 66%; (h) E . subtilis fraction VIII; enzyme recovery 90%; protein recovery 105%.
1686
with a linear gradient between 50 and 250 m~ potassium phosphate, pH 8.0, the total volume of the eluent being 500 ml. The peak fractions were pooled (fraction VIII). Conentration and storage at 0" were as described for the E. coli enzyme. The purification can also be carried out successfully by reversing steps VI and VII, with appropriate modification of the volume of the columns and gradients. C. Charucterization of Purified Enzyme Preparations. I. ELECTROPHORETIC BEHAVIOR. Electrophoresis of the E . subtilis tyrosyl RNA synthetase on starch gel (Figure 1) or polyacrylamide gel (Figure 2) revealed primarily one protein-containing band which coincided with the position of the enzymatic activity. The enzyme from E. coli contained additional minor protein bands (Figure I), hut in both the starch and polyacrylamide gel separations the activity migrated only with the major component. Photodensitometer tracings of the photographs shown in Figure 1 and integration of the protein peaks indicate that the maximum purities of the E. subfilis (fraction VIII) and E. coli (fraction MI)preparations are 90 and 70 %, respectively. Electrophoresis of a mixture of the two tyrosyl RNA synthetases on starch gel (Figure 1) or polyacrylamide gel (Figure 2) shows that the two enzymes are readily separated; the E. subfilis enzyme is clearly more anionic at the pH values of 8 and 10 used in the two procedures. 11. SLDIMENTATION BEHAVIOR AM) MOLECULAR SIZE. Sedimentation of the E . subtilis and E. coli tyrosyl RNA synthetases through a sucrose gradient showed that in both cases the enzymatic activity and the protein boundary sediment4 together (Figure 3a and 3b). Lighter and heavier impurities w r e noted in the E. coli
RICHARD C A L E N D A R A N D PAUL BERG
4 : Schlieren photographs of sedimentation patterns of purified tyrosyl RNA synthetases. (a) E. coli enzyme at\7.5mg/ml. The rotor speed was 50,740 rpm at 6.2' with a schlieren angle of 75". Times of photography were 32, 48, and 68 min. (b) E. subfilis enzyme at 6 mpiml. The rotor speed was 50,740 rpm at 6.7" with the schlieren angle of 75". Times of photography were 36,56, and 80 min. FIGUR~
enzyme (fraction VII), and the specific activity in fractions 21-29 varied from 1120 to 1520 units/mg. The E. subrilis enzyme (fraction VIII) showed an enzyme peak of nearly constant specific activity (5WM)O unitslmg), although there was a small amount of lighter protein. If the specific activity at the peak is taken as the specific activity of pure enzyme, the upper limits of the purities of the E. coli (fraction VII) and E. subrilis (fraction VIII) enzymes are 89 and 94%. respectively. The coincidence of protein and enzymatic activity in these sedimentation studies suggest that these enzyme preparations are suitable for the determination of the enzyme molecular weights in the analytical ultracentrifuge. Measurement of the sedimentation coefficients of the two enzymes were also made at several protein concentrations using schlieren optics. Figure 4 (a and h) shows the Schlieren pattern for the E. coli and E. subtilis enzymes and Figure 5 shows the dependence of snO,- on protein concentration. Extrapolation to zero ~ . of ~5.2 and protein concentration yielded s ~ values 5.2 for the E. coli and E. subtilis tyrosyl RNA syuthetases, respectively. Little or no aggregation is seen with the E. subtilis enzyme, but there is less certainty of this with the E. colienzyme. The diffusion coefficient, D20,,,was determined in one experiment as 4.7 X l e 7 cmz/sec for the E. coli tyrosyl RNA synthetase, at an initial concentration of 6.0 mglrnl. The value of Dlo.wfor the E. subrilis enzyme was measured in three different experiments as 5.3, 5.2,
VOL.
5,
NO.
5,
~
TABLE 111:
MAY
~~
1966
~
Molecular Weight Values of Tyrosyl RNA Synthetases from E. coli and B. subtilis:
Enzyme a. E. coli tyrosyl RNA synthetase b. B. subtilis tyrosyl RNA synthetase
Protein Concn (ms/ml)
Rotor Speed (rpm)
3.7 5.0 6.0 6.0 3.0
19,160 17,250 9,341 15,220 8,225
M,
MZ
M*
85,000
95,000 98,000 90,000 92,000 88,000
95,000 96,000 86,000 91,000
a Equilibrium centrifugation results. M,, and M , were determined using method I and method I1 (Van Holde and Baldwin, 1958) as described earlier. M* was obtained from the slope of a plot of In c cs. rz(Schachman, 1957).
~~
Amino Acid Composition of Tyrosyl RNA Synthetases of E . coli and B. subtilis. TABLE IV:
Approx Residues/Mole Amino Acid Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Lysine Histidine Arginine Tryptophan Half-cystine Total
+ asparagine
+ glutamine
E. coli
B. subtilis
90 40 38 111 37 79 77 51 18 51 84 26 41 62 13 45 $18 15
91 48 50 102 21 61 55 39 10 54 85 31 39 64 10 39 14 2-3
896
81 5
and 5.8 X 10-7 cm2/secat 6.0, 3.8, and 3.0 mg of protein/ml, respectively. When s ~ ~ and ., DzOjware combined in the Svedberg equation (Svedberg and Pedersen, 1940), molecular weight values of 97,000 and 87,000 are found for the E. coli and B. subtilis enzymes, respectively. Sedimentation equilibrium studies were carried out using short columns of solution (1.5-3 mm). Plots of l/r(dc/dr) us. ( c - c,) obtained from measurements of Schlieren photographs (method 11, Van Holde and Baldwin, 1958) are shown in Figure 6. The fact that the plots are straight lines indicates reasonable homogeneity and lack of aggregation. The values determined for
C H A K A C T E K I Z A I I O N
4
*-
e Protei" m9
p e r rnl
FIGURE 5 : The dependence of sedimentation coefficient on protein concentration. The conditions of sedimentation are given in the legend to Figure 4 and in Methods.
molecular weight are given in Table 111; the agreement of M , and M, further indicates the homogeneity of these preparations. The E. coli enzyme has a molecular weight of about 96,000, while the B. subtilis enzyme is ca. 88,000. 111. AMINOACID COMPOSITION. Table IV presents the results of amino acid analyses carried out on the tyrosyl RNA synthetases of E . coli and B. subtilis. The values are presented as residues per molecule of enzyme. The most distinctive difference in amino acid composition of these two enzymes is their half-cystine content. The enzyme from E . coli appears to have 15 half-cystine residues, while the B. subtilis enzyme has only two or three. It is possible that some or all of the cysteic acid in the E. coli preparation comes from contaminating proteins. The value for two half-cystines/mole for the B. subtilis enzyme is not due to incomplete oxidation of cysteine and cystine, since the value for methionine (determined in the oxidized sample as methionine sulfone) is the same for oxidized and unoxidized samples of the protein. The partial specific volumes estimated from the amino acid composition (Cohn and Edsall, 1943) were 0.730 for the E. coli enzyme and 0.735 for the B. subtilis enzyme. IV. IMMUNOLOGIC CHARACTERIZATION OF E. coli AND
OF
r Y R o s Y L
RIBONUCLEIC
ACID
1687
SYNTHETASE
UIOCHBMISTRY
1958) of sedimentation equilibrum data for purified tyrosyl RNA synthetases. A. E. coli enzyme at 3.7 mg/ ml initial concentration. The rotor speed was 19,160 rpm for 33 hr at 6.2". The column height was 1.88 cm. B. B. subtilis enzyme at 6.0 mgjml initial concentration. The rotor speed was 9341 rpm for 53 hr at 6.7'. The column height was 3.25 cm.
B. suhrilis TYROSYL SYNTHErAsw. The E. coli (fraction V111) and B. subtilis (fraction VIII) tyrosyl RNA synthe-
1688
tases were tested also for immunologic cross-reaction. Figure 7 shows that antiserum prepared against the E. coli enzyme (AE) formed a single precipitin line with the E. coli enzyme but no detectable precipitate with the B. suhtilis enzyme. Antiserum to the B. suhtilis enzyme precipitated the B. subtilis enzyme and gave a reaction of partial identity with E. coli enzyme. This latter cross-reaction was seen after acid fixation of the gel by the faint precipitin band between the wells marked E and AB (Figure 7,B). A more quantitative estimate of the immunologic cross-reactivity was obtained by studying neutralization of enzyme activity by y-globulin isolated from the antisera. y-Globulin directed against the purified E. coli enzyme could neutralize (>99 %) the tyrosine-dependent A T P - T P j exchange activity of purified E. coli enzyme or of an E. coli extract; however, there was no significant inhibition of activity of the purified B. suhtilis enzyme when compared to a normal nonimmune y-globulin (Figure SA). With y-globulin from animals immunized with the B. suhtilis enzyme, the B. subtilis enzyme (purified or crude) was neutralized, but there was definite indication of some neutralization of the heterologous E. coli enzyme (Figure 8B). The heterologous reaction was, however, only c a 10% that of the homologous reaction. Neither immune y-globulin preparation neutralized the activity of purified isoleucyl RNA synthetase (Baldwin and Berg, 1966). Thus, by immunologic criteria, as well as electrophoretic be-
KICHAKII
CALENDAR A N D
13AUL H B K G
FIGURE l : Initnunodithion using antisera against purified tyrosyl RNA synthetases. 1-lie holes contained 0.1 ml of solution as follows: (AE) undiluted antiserum to E. coli enzyme; (E) E. coli enzyme (fraction VIII), 1.8 mgiml; (B) B. subtilis enzyme (fraction VIII), 1.5 mgjml; (AB) undiluted antiserum to B. suhtilis enzyme; (a) before acid fixing; (b) after acid fixing.
havior, the tyrosyl RNA synthetase from E. coli and B. subtilis are readily distinguishable. Discussion Although the two tyrosyl RNA synthetases catalyze the identical reaction and can be used interchangeably with t-RNA from E. coli or B. subtilis (Calendar and Berg, 1966) they are distinguishable by several criteria. The two enzymes have different electrophoretic mobilities in starch or polyacrylamide gels, the B. subtilis enzyme being more anionic. Of considerable interest is the striking difference in the half-cystine content of the two enzymes. The E. coli enzyme contains 15 halfcystine residuesjmole, an amount similar to that found for E. coli isoleucyl RNA synthetase (Baldwin and Berg, 1966), while the B. subtilis enzyme has two or three. The tyrosyl RNA synthetases from E. coli and B. subtilis are also differentiated by immunologic criteria. Anti-E. coli tyrosyl RNA synthetase precipitates and completely neutralizes the E. coli enzyme but has no effect on the B. subtilis activity; anti-B. suhti1i.s tyrosyl RNA synthetase precipitates and neutralizes the homologous enzyme but cross-reacts with the E. coli enzyme. The analysis of this partial cross-reaction in terms of the two protein structures needs further study. Based upon the amount of purified protein recovered in each purification procedure one cannot determine if only a single tyrosyl RNA synthetase exists in the organisms examined. However, the neutralization of 95-100z of the tyrosyl RNA synthetase activity of crude extracts by antisera prepared against the most
VOL.
5, NO. 5, M A Y 1 9 6 6 Anti- EColi Enzyma Gamma Globulin
Anti 6 . S u b t i l i s Enzyme Gomma Globulin /I I I I
I
0 0
0
6 Sub-
\
0
\ E
C&
100
ii
0
0
Enzyme
Enzyme
200
300 0
500
p q r a m s qamma q l o b u l i n
I000
1500
8: Neutralization of tyrosine-dependent ATP-32PPi exchange activity with immune y-globulin. Mixtures of 0.33 units of enzyme (0.6 pg of protein) and antiserum in 0.20 mM potassium phosphate, pH 6.5, 1 mM 2-mercapto. ethanol, 250 pg/ml of bovine plasma albumin, were incubated at 37" for 5 min, kept at room temperature for 45 min and stored at 4" overnight before centrifugation at 13,OOOg for 20 min. A. Reaction of y-globulin directed against purified E. coli tyrosyl RNA synthetase with: purified B. subtilis enzyme; purified E . coli enzyme; E. coli dialyzed autolyzed extract (Table I, fraction 11). B. Reaction of y-globulin directed against purified B. subtilis tyrosyl RNA synthetase with: purified E. coli enzyme; purified B. subtilis enzyme; B. suhtilis dialyzed, autolyzed extract (Table 11, fraction TI).
FIGURE
purified preparations (Figure 8) suggests that the enzymes isolated are probably the exclusive species present in the cell. Any multiple forms of the enzyme would have to be extremely labile or immunologically crossreacting. If the proteins isolated to represent the total complement of these enzymes in the cell, then there are ca. 1400 and 2000 molecules of enzyme/cell in E. coli and B. subtilis, re~pectively.~ This is of the same order as that necessary to assure generation times of 19 and 40 rnin for E. coli and B. subtilis, respectively.*It is also of interest that based on the yield of t-RNA,, obtained from these bacteria grown under the same conditions 7 1012 E. coli cells (1-g wet weight) contain 150 units of enzyme; pure enzyme contains 7.0 X 1016 enzyme molecule (750 units)/mg of enzyme. Therefore, if enzyme molecules in the crude and purified states have the same activity, then 1012 cells contain 1.4 X 1015 enzyme molecules or 1400/cell. The pertinent values for B. subtilis are: 1011 B. subtilis cells (1 g wet weight) contain 20 units of enzyme; pure enzyme contains 6.5 X loL5 enzyme molecules (750 units)/mg of enzyme. Therefore, 1011 cells contain approximately 1.7 X 10" enzyme molecules or 1700/cell. 8 10'2 E. coli cells (1-g wet weight) contain 150 units of enzyme. One unit of enzyme will esterify 0.05 pmole of tyrosine to t-RNA (E. coli) in 15 min at 37"; 9.6 pmoles of L-tyrosine are contained in the protein of 1 wet g of E. coli (Sueoka, 1961; Long, 1961), and this amount of tyrosine can be esterified to t-RNA in 19 rnin by the tyrosyl R N A synthetase of a crude E. coli extract. 1011 B. subtilis cells (1-g wet weight) contain 20 units of enzyme; one unit of enzyme can esterify 0.15 pmole of tyrosine to t-RNA ( E . coli). L-Tyrosine, 8 pmoles, is contained in the protein of 1 wet g of B. subtilis (Sueoka, 1961; Long, 1961), and this amount of tyrosine can be esterified to t-RNA in 40 rnin by the tyrosyl R N A synthetase of a crude B. subtilis extract. In these calculations it is assumed that enzyme in the intact cell has the same activity as that in a crude extract.
CHARACTERIZATION
(R. Calendar, unpublished data, 1966) there are cu. 1000 molecules of t-RNA,,,/E. coli cell and 5000/B. subtilis celI.9 Thus, it may be that the ratio of enzyme to t-RNA is near 1. References Baldwin, A. N., and Berg, P. (1966), J. Biol. Chem. 241, 831. Beaven, G. H., and Holiday, E. R. (1952), Advan. Protein Chem. 7, 319. Benzer, S., and Weisblum, R. (1961), Proc. Natl. Acad. Sci. U. S.47, 1149. Berg, P. (1961), Ann. Rev. Biochem. 30, 293. Bergmann, F., Berg, P., and Dieckmann, M. (1961), J. Biol. Chem. 236, 1735. Calendar, R., and Berg, P. (1966), Biochemistry 5,1690. Cohn, E. J., and Edsall, J. T. (1943), Proteins, Amino Acids and Peptides, New York, N. Y.,Reinhold, p 375. Doctor, B. P.,and Mudd, J. A. (1963), J. Biol. Chem. 238,3617. Jovin, T., Chrambach, A., and Naughton, M. A. (1964), Anal. Biochem. 9, 351. Long, C. (1961), Biochemists' Handbook, Princeton, N. J., D. Van Nostrand Co. Lowry, O., Rosebrough, N., Farr, N., and Randall, R.
9 E. coli cells (1-g wet weight) contain 1.7 mpmoles of t-RNAt,,, Le., 1020 molecules of t-RNAt,/E. coli cell; 10" B. subtilis cells (1-g wet weight) contain 0.8 mpmole of t-RNAtYr, Le., 4800 molecules of t-RNAt,,/B. subtilis cell. In these calculations it is assumed that assays using purified tyrosyl R N A synthetase and whole t-RNA (Calendar and Berg, 1966) define the correct amount of tyrosine acceptance in the cell.
OF T Y R O S Y L R I B O N U C L E I C
ACID
1689
SYNTHETASE
BIOCHEMISTRY
(1951), J. Biol. Chem. 193, 265. Martin, R., and Ames, B. (1961), J. Biol. Chem. 236, 1372. Moore, S . (1963), J. Biol. Chem. 238, 235. Nester, E. W., Schafer, M., and Lederberg, J . (1963), Genetics 48, 530. Ouchterlony, 0. (1948), Acta Pathol. Microbiol. Scand. 25,186. Richardson, C. C., and Kornberg, A. (1964), J. Biol. Chem. 239, 242. Sanger, F., and Thompson. E. 0. P. (1963), Biochim. Biophys. Acta 71, 468. Schachman, H. K. (1957), Methods Enzymol. 4, 32. Smithies, 0. (1959), Biochem. J. 71, 585.
Spackman, D. H., Stein, W. H., and Moore, S. (1958), Anal. Chem. 30, 1190. Spizizen, J. (1958), Proc. Natl. Acad. Sci. U.S . 44, 1072. Sueoka, N. (1961), Proc. Natl. Acad. Sci. U.S. 47, 1141. Svedberg, T., and Pedersen, K. 0. (1940), The Ultracentrifuge, Oxford Press, New York, N. Y. Van Holde, K. W., and Baldwin, R. L. (1958), J. Phys. Chem. 62, 734. Weismeyer, H., and Cohn, M. (1960), Biochim. Biophys. Acta 39, 417. Willstatter, R., and Kraut, H. (1923), Ber. Deut. Chem. Ges. 56, 1117. Yamane, T., and Sueoka, N. (1963), Proc. Natl. Acad. Sci. U. S. 50, 1093.
The Catalytic Properties of Tyrosyl Ribonucleic Acid Synthetases from Escherichia coli and Bacillus subtilis" Richard Calendar and Paul Berg
ABSTRACT: A comparison of the substrate specificity of tyrosyl RNA synthetases from Escherichia coli and Bacillus subtilis has been made. Of the four common ribonucleoside triphosphates only adenosine triphosphate is utilized by both enzymes. Several alterations in the L-tyrosine structure have been examined for their effect on the formation of aminoacyl adenylate. Removal of the carboxyl group (tyramine) or its reduction (L-tyrosinol), amidation (L-tyrosine amide), or esterification @-tyrosine methyl ester) yielded inactive substrates although each of these analogs was a competitive inhibitor. Modification of the a-amino group (N-acetyl or N-glycyl) or substitution of the a hydrogen by a methyl group yielded compounds which were neither substrates nor inhibitors. Quite unexpectedly, Dtyrosine was activated and transferred to t-RNA,,, by
I
n the previous paper (Calendar and Berg, 1966) tyrosyl RNA' synthetases from E . coli and B. subtilis were isolated and several of their physical and chemical properties were compared. These two enzymes have
both enzymes. But in each case for the exchange reaction the V,, was lower and the K , was higher than with Ltyrosine. A shift of the ring hydroxyl group from the 4 (para) to the 3 (meta) or 2 (ortho) position yielded inactive substrates with both enzymes. Although substitution of chloro, iodo, amino, or nitro groups in position 3 of L-tyrosine eliminated the substrate activity, after introduction of a fluoro or hydroxy group in the same position the derivatives were still active, but the VmSx was lower and the K , was higher. Conversion of the benzene ring to a pyridine structure (5-hydroxy-2-(3-~~alany1)pyridine) decreased the Vmx to about half and increased the K , about 30-fold with both enzymes. The B. subtilis and E . coli tyrosyl RNA synthetases utilize the t-RNAty, from either bacterial source; the same amount of either tyrosyl RNA is formed with either enzyme.
also been examined with respect to several catalytic parameters, particularly those related to the specificity for the substrates and analogs of the substrates, and these results are reported here. Experimental Section
* From
1690
the Department of Biochemistry, Stanford University School of Medicine, Palo Alto, California. Receiced February 15, 1966. This work was supported in part by teaching and research grants from the National Institutes of Health. 1 Abbreviations used: RNA, ribonucleic acid; t-RNA, transfer R N A ; F-tyrosine, 3-fluorotyrosine; DOPA, 3-hydroxytyrosine; HPA, 5-hydroxy-2-(3-~~-alanyl)pyridine; ATP, GTP, CTP, and UTP, adenosine, guanosine, cytidine, and uridine triphosphate; ADP, adenosine diphosphate; dATP, deoxy ATP; dTTP, deoxythymidine triphosphate.
RICHARD CALENDAR
AND
PAUL
B E R G
Materials. Glycyl-L-tyrosine, N-acetyl-L-tyrosine, 3hydroxyl-L-tyrosine, and natural L-amino acids were purchased from California Corp. for Biochemical Research. D-Tyrosine, 3-iodo-~-tyrosine, 2-hydroxy-~3-amino-~phenylalanine, 3-hydroxy-~-phenylalanine, tyrosine, 3-hydroxy-~~-tyrosine.and 4-fluoro-DLphenylalanine were purchased from Nutritional Biochemicals, Inc., Cleveland, Ohio. 4-Chloro-DL-phenyl-
